fastxt.app

What It Is

Fastxt is a cross-platform text and note management tool built around a “local-first” philosophy and on-device AI. It stores text in a local SQLite database, does not rely on a centralized service, and supports multi-device sync over the same Wi‑Fi/LAN via QR code scanning. Its AI features also emphasize local processing, aiming to organize, search, and understand notes without uploading note content.

Core Features and Models

Fastxt’s AI capabilities include smart tagging, long-form summarization, semantic search, and AI-assisted classification. On desktop, it relies on Ollama and allows users to configure a local endpoint and model name. llama3.2 is recommended by default, with options such as llama3.2:1b, mistral, and phi3 also listed. On iOS/macOS, it uses Apple Foundation Models, but requires Apple Silicon, iOS 18.1+ or macOS 15.1+, and Siri must be enabled. On Android, it uses on-device ML capabilities and requires Android 14+ plus sufficient memory.

Pricing, Privacy, and Integrations

The crawled content does not disclose pricing, free quotas, or payment methods, so the commercial cost of using it remains unclear. Privacy is its biggest highlight: the documentation explicitly states that notes do not leave the device, there is no cloud processing, and no data is collected or transmitted. Ollama also does not call external APIs. Advanced users can additionally invoke features such as ai-tag, ai-summarize, ai-embed, semantic-search, and ai-organize through the command interface, making it suitable for users with some technical background who want to build automation workflows.

Pros, Cons, and Best-Fit Users

Its strengths are clear privacy boundaries, offline availability, AI features that align well with note organization scenarios, and multi-platform support. Its limitations are that AI quality depends on local models and hardware; lower-end devices may be slow or run out of memory; mobile AI has OS and chip requirements; sync is LAN-oriented, with no clear cloud sync option; and the Chinese interface and Chinese-language AI performance are not clearly described. It is best suited for privacy-conscious personal knowledge management users, developers, researchers, and local AI users who are comfortable using Ollama.

Access in China and Alternatives

The main content does not provide information about access from mainland China, network connectivity, or payment support, so china_access can only be assessed as unknown. Because the desktop version depends on Ollama, local model downloads and the runtime environment may affect the actual experience. Comparable alternatives include Obsidian, Joplin, Logseq, Apple Notes, and self-hosted Ollama + local note-taking workflows.